Home-made Vactrols - trials and tribulations

[jrc4558]

Here's how I do it.
1) start with a super-bright clear 5mm LEDs (red of course, they're cheaper)
2) sand the top of the LED off, to turn it into a cylinder.
3) glue it with a tiny drop of superglue on top of an LDR
4) roll it one layer of a utility aluminum foil (the one you use for baking) shiny side inwards)
5) heatshrink
6) one more layer of heatshrink.
The result - well, a good optocoupler. I match the LDRs approximately in the same way as RG suggests, except i came up with a slightly different gadget that essentially is a little plastic tube which allows to push in any LED and any LDR (under 8 mm in dia) in on either end and is shiny inside.
Good luck to you all.

[DiyFreaque]

The range went from infinite to pretty close to a dead short (at least in the eyes of the PT2399), IIRC.  I had pretty much the same range as a pot tied to the clock timing resistor pin, so you could instantly move the PT2399 through-out the entire range of delay, set by the pot on the timing pin (read on).

I used an op amp to mix the CV with an adjustable negative offset, around -2.1V to 0V, and that control could be used to set the inital delay. The CV could be either bipolar or unipolar.  The CV was scaled so that 10Vp-p (synth levels) were reduced to 2Vp-p max to operate within that range, with an input attenuator for reducing the CV's effect to smaller deviations.

The PT2399 doesn't sound terribly good (or at all) with an infinite resistance as the timing resistor, so I also had a 100K pot in parallel with the FET (100K gets the longer albeit grungier echoes out the PT2399, a 50K puts it in the 'normal' range of the PT2399), so the max resistance at that pin would be 100K.  

So, I ended up with two different controls over the delay time - the 'inital delay' CV control, and the 100K pot.  With this setup, one could control the range that the CV operated in with the 100K pot - say reducing it to 25K would allow the CV to move from the delay time the PT2399 would be at with 25K resistance at the pin to shorter delay times.  That was handy for fine-tuning very small, short range delays with the CV controls.

It was no Volt/Octave type of control scheme, but then again, I don't think the linearity was any worse than using an optocoupler and it was a helluva lot faster and got to the smaller delays much, much better.  A sample and hold input, at very small minute, barely perceptible deviation, with a longish echo with a high number of repeats produced a really nice spacy effect.  Higher levels of deviation using a sample and hold produced those bleepy computer-like sound effects that Sci-Fi movies in the 50's and 60's always associated with computers of the future....

Cheers,
Scott

[Arno van der Heijden]

I was thinking of using this approach in the Vanishing Point (http://www.geocities.com/thetonegod/vanpoint2/van_point_schematic.gif), where it should replace one of the LED/LDR combo's. However, the optocoupler is used here in the feedback loop of a schmitt inverter based oscillator. It is set up so that one oscillator varies the frequency of the second one.

Therefore my question: can it be used in this configuration?

Right now, I'm trying with a normal transistor, as described in this article: http://www.geofex.com/Article_Folders/LFOs/psuedorandom.htm

[Arno van der Heijden]

bump

[DiyFreaque]

Hi Arno,

Cool stuff BTW - both the Vanishing Point and R.G.'s LFO page - how did I miss that?

Anyway, I'd say that a FET would work in that application as a VCR, though it might take a bit of circuitry addition and tweaking to drive the FET.  In this case, I'm not sure what advantage would be gained controlling the oscillator with a FET VS the Optocoupler.  Another factor to consider is that the properties of the Vanishing Point may rely on the slower reactance of the LDR (I don't know).  

Best thing to try would be to breadboard it and see if it flies!

Take care,
Scott

[Joe Kramer]

Hi!

Mouser's p/n 606-CMD5777C is a diffused red LED with a viewing angle of 180 degrees.  I've used this in the "sitting around the campfire" implementaion with four LDRs for a univibe-type effect with great success.

The trick to getting the LED and LDRs in harmony was to use a 10-turn trimmer connected to the V+ and the LED to inject a small bias current.  Once the LED was set to it's "just on" point, the whole circuit came to life.  The importance of the LED in the equation can't be underestimated.

Hey Mark: don't give up on those DIY vactrols.  Back in the day, I once made up a batch of my own, thinking I'd beat the cost of CLM6000s.  They tested and worked great until I sealed them up in epoxy--epoxy impregnated with aluminum, and hence CONDUCTIVE!  Shot in the foot by his own gun!  But now I'm back at it, and the limp is hardly noticable. . . .

Joe

[The Tone God]

Sorry for the late reply folks. I didn't see this thread untill I was doing a search for something.

Cool stuff BTW - both the Vanishing Point and R.G.'s LFO page - how did I miss that?

Thanks.

Anyway, I'd say that a FET would work in that application as a VCR, though it might take a bit of circuitry addition and tweaking to drive the FET.  In this case, I'm not sure what advantage would be gained controlling the oscillator with a FET VS the Optocoupler.

I don't see a major advantage either considering all the extra circuitry that would be needed. My concerns with using a FET in the VP is possible bias issues and the possible range that the FET could swing through. That would make the VP really painful to tune especially over a wide voltage range. The optocoupler to me just seemed to be the easy and safe way to get things working. K.I.S.S.

Another factor to consider is that the properties of the Vanishing Point may rely on the slower reactance of the LDR (I don't know).

Nah the VP doesn't really care about "swing" charateristics in the high frequwncy clock (HF clock). Its there just to adjust the HF clock through a wide range to reduce the likelyness of hitting the same number of clock pulse in a row. How it swings through the frequency range is not that important.

For the audio interface the response of the optocoupler does come in handy as it reduces the "sharpness" of the stage switching. The Original Vanishing Point was fairly stiff in that respect which added alittle ticky-ness when switch between extreme settings. The optocoupler response soften that.

Also using the optocoupler helped isolate the audio from the digital nicely. With a FET you don't get that isolation so noise could be a major problem.

Andrew

[DiyFreaque]

Nah the VP doesn't really care about "swing" charateristics in the high frequwncy clock (HF clock). Its there just to adjust the HF clock through a wide range to reduce the likelyness of hitting the same number of clock pulse in a row. How it swings through the frequency range is not that important.

I couldn't see why the response of the optocoupler would be crucial to the operation, but then again, I always have my doubts about those things until I actually try them.  Every now and then, it turns out the part the designer used is the part you better use if you want the same results  

For the audio interface the response of the optocoupler does come in handy as it reduces the "sharpness" of the stage switching. The Original Vanishing Point was fairly stiff in that respect which added alittle ticky-ness when switch between extreme settings. The optocoupler response soften that.

Optocouplers are great for that, aren't they?  One design that always intrigued me was Grant Richter's exquisite corpse thing, which takes advantage of the response time of pulse modulated Vactrols to randomly crossfade between up to eight inputs, using a bargraph driver (two modes - dot and bar!).

If you had an effects chain with several effects running in parallel, this thing (when controlled with a smooth random voltage) would randomly fade from one line to the next - for example, fade from distorted, to clean, to flange to phase, etc, in the 'dot' mode.  If one had a pedal to put out the CV, you could control it that way, too.

http://www.musicsynthesizer.com/Corpse/Exquisite%20Corpse2.html

Take care,
Scott

[Mark Hammer]

I succumbed over the weekend, and built up the envelope follower from the Bi-Filter Follower on its own board, with attenuator pot at the input and bias-voltage adjust, as well as solid core leads to connect up to the LED half on a breadboard.  Certainly nice for testing out candidate LEDs (always good to have a sense of how bright that little sucker glows in response to wacka-wacka strumming), although I suspect that it could benefit from having some means to either adjust the gain of the envelope follower stage in designated increments, or some input stage prior to the sensitivity control with known gain increments.